Environmental Modeling Packages for the MSTDCL TDP

Transcription

1 Copy No. Defence Research and Development Canada Recherche et développement pour la défense Canada Environmental Modeling Packages for the MSTDCL TDP Review and Recommendations Robert Trider, Peter Giles and Bruce Martin General Dynamics Canada Ltd. General Dynamics Canada Ltd Richmond Road Ottawa, Ontario K2H 5B7 GDC Document Number: Contract Project Manager: Beatrice Chappaz, Contract Number: W Contract Scientific Authority: Dr. Mark Trevorrow, ext 315 The scientific or technical validity of this Contract Report is entirely the responsibility of the Contractor and the contents do not necessarily have the approval or endorsement of Defence R&D Canada. Defence R&D Canada Atlantic Contract Report DRDC Atlantic CR September 2009

2 This page intentionally left blank.

3 Environmental Modeling Packages for the MSTDCL TDP Review and Recommendations Robert Trider Peter Giles Bruce Martin General Dynamics Canada Ltd General Dynamics Canada Ltd 3785 Richmond Rd. Ottawa, Ontario K2H 5B7 GDC Document Number: Contract Project Manager: Beatrice Chappaz, Contract Number: W CSA: Dr. Mark Trevorrow, Scientific Advisor MSTDCL TDP, x315 The scientific or technical validity of this Contract Report is entirely the responsibility of the Contractor and the contents do not necessarily have the approval or endorsement of Defence R&D Canada. Defence R&D Canada Atlantic Contract Report DRDC Atlantic CR September 2009

4 Principal Author Peter Giles General Dynamics Canada Approved by Original signed by David Hazen David Hazen Head, Technology Demonstration Approved for release by Original signed by Ron Kuwahara for Dr. Calvin Hyatt Head, Document Review Panel Her Majesty the Queen in Right of Canada, as represented by the Minister of National Defence, 2009 Sa Majesté la Reine (en droit du Canada), telle que représentée par le ministre de la Défense nationale, 2009

5 Abstract.. Since 2005 DRDC Atlantic has been conducting the Multi-Sensor Torpedo Detection, Classification, and Localization (MSTDCL) technology demonstration project aimed at improving the torpedo detection, classification, and tracking capabilities on Halifax-class frigates. This document examines the advantages to the MSTDCL project of adding a capable Environmental Analysis package for detection performance prediction. Three levels of complexity were examined: a basic level based on the Networked Underwater Warfare (NUW) developed analysis package, an intermediate level package building on the NUW package to provide improved functionality and displays while reducing operator interaction, and an Advanced Environmental Analysis package that improves the accuracy of the performance predictions by more accurately representing range-dependent environments. The advantages of each level to the MSTDCL system are compared, along with estimates of the work level required to implement the package. A low-risk approach beginning with the NUW package and advancing through the intermediate levels is recommended. Résumé... Depuis 2005, RDDC Atlantique mène le projet de démonstration de la technologie de détection, de classification et de localisation des torpilles à partir de capteurs multiples (DCLTCM), qui vise l'amélioration des capacités de détection, de classification et de poursuite des torpilles à bord des frégates de la classe Halifax. Le présent document examine les avantages, pour le projet DCLTCM, de l'ajout d'une trousse satisfaisante d'analyse environnementale en vue de la prédiction du rendement de la détection. Trois niveaux de complexité ont été examinés : un niveau de base, fondé sur la trousse d'analyse élaborée dans le cadre de la guerre sous-marine en réseau (GSR); une trousse intermédiaire, qui fait fond sur la trousse GSR pour permettre l'amélioration de la fonctionnalité et les affichages, tout en réduisant l'interaction de l'opérateur; et une trousse d'analyse environnementale avancée, qui permet d'améliorer la précision des prédictions du rendement grâce à une représentation plus précise des milieux dépendants de la distance. Les avantages de chaque niveau pour le système DCLTCM sont comparés, et le niveau de travail requis pour la mise en œuvre de la trousse est estimé. On recommande une approche à faibles risques, qui commence par la trousse GSR et se poursuit par les niveaux intermédiaires. DRDC Atlantic CR i

6 This page intentionally left blank. ii DRDC Atlantic CR

7 Executive summary Environmental Modeling Packages for the MSTDCL TDP: Review and Recommendations Robert Trider; Peter Giles; Bruce Martin; DRDC Atlantic CR ; Defence R&D Canada Atlantic; September Introduction: Since 2005 DRDC Atlantic has been conducting the Multi-Sensor Torpedo Detection, Classification, and Localization (MSTDCL) Technology Demonstration Project (TDP) aimed at improving the torpedo detection, classification, and tracking capabilities on Halifaxclass frigates. The basic thrust has been the integration and display of sonar data from the SQR-19 towed array, the SQS-510 sonar, sonobuoys, and the ship s command and control system, resulting in a sea-going system known as PLEIADES. As part of MSTDCL the feasibility of developing a near-real-time underwater acoustic sensor performance prediction capability was explored. Previously, the Networked Underwater Warfare (NUW) project had developed an underwater environmental modeling package, based on earlier work at DRDC Atlantic. Results: The capabilities provided by the NUW acoustic modeling system define a basic environmental modeling package for the purposes of this study. This study reviews this basic package, discusses its potential advantages to the MSTDCL TDP and PLEIADES system, and then considers expanded capabilities. Specific suggestions are made for improving this basic package by adding improved environmental databases and tools, improving the prediction accuracy through addition of a parabolic-equation model for low-frequencies, and improving operator tools. This study discusses some shortcomings of the basic package for littoral operations, and then reviews some of the latest adaptive modeling techniques which could be employed. A graduated approach beginning with the basic NUW capability is recommended. Significance: The performance of underwater acoustic sensors is heavily dependent on local environmental conditions, especially in littoral regions, thus it is essential to be able to predict sensor performance in near real time. Considerable work on development of accurate underwater acoustic performance models has been done by DRDC in the past decades. This study proposes a graduated suite of environmental prediction capabilities that should be implemented within the PLEIADES sonar processing system. Ultimately this will provide a sophisticated sonar performance modeling capability that is integrated with the sonar processing and automated detection and classification functions. Future plans: This study was completed in late-2007, and since then work to integrate the basic NUW sonar performance capability into the PLEIADES system has been performed; specifically implemented on HMCS Ottawa during Westploy and RIMPAC exercises in mid Continued development of environmental modeling capabilities at DRDC Atlantic is being conducted under project 11ch Assessing Sonar Performance in a Realistic Environment. DRDC Atlantic CR iii

8 Sommaire... Environmental Modeling Packages for the MSTDCL TDP: Review and Recommendations [Trousses de modélisation environnementale pour le PDT DCLTCM: Revue et Recommendations] Robert Trider; Peter Giles; Bruce Martin; DRDC Atlantic CR ; R & D pour la défense Canada Atlantique; Septembre Introduction : Depuis 2005, RDDC Atlantique mène le projet de démonstration de la technologie (PDT) de détection, de classification et de localisation des torpilles à partir de capteurs multiples (DCLTCM), qui vise l'amélioration des capacités de détection, de classification et de poursuite des torpilles à bord des frégates de la classe Halifax. Le projet visait essentiellement l'intégration et l'affichage des données du sonar remorqué SQR-19, du sonar SQS-510, de bouées acoustiques et du système de commandement et de contrôle du navire, ce qui a donné un système de mer appelé PLEIADES. Dans le cadre du DCLTCM, on a étudié la faisabilité de la mise au point d'une capacité de prédiction du rendement de capteurs acoustiques sous-marins en temps quasi réel. Plus tôt, dans le cadre du projet de guerre sous-marine en réseau (GSR), on avait mis au point une trousse de modélisation environnementale sous-marine, d'après des travaux antérieurs menés par RDDC Atlantique. Résultats : Les capacités fournies par le système de modélisation acoustique GSR définissent une trousse de modélisation environnementale de base aux fins de la présente étude. Dans la présente étude, on examine la trousse de base, on discute de ses avantages possibles pour le PDT DCLTCM et le système PLEIADES, puis on considère les capacités élargies. Des propositions précises sont faites en vue de l'amélioration de la trousse de base par l'ajout de bases de données environnementales et d'outils améliorés, l'amélioration de la précision des prédictions par l'ajout d'un modèle d'équation parabolique pour les basses fréquences, puis l'amélioration des outils de l'opérateur. Dans la présente étude, on discute de certaines lacunes de la trousse de base pour les opérations littorales, puis on examine certaines des dernières techniques de modélisation qui pourraient être employées. On recommande une approche progressive qui commence par la capacité GSR de base. Portée : Le rendement des capteurs acoustiques sous-marins dépend énormément des conditions ambiantes locales, en particulier dans les régions littorales. C'est pourquoi il est essentiel d'être en mesure de prédire le rendement des capteurs en temps quasi réel. Au cours des dernières décennies, RDDC a mené beaucoup de travaux en vue de la mise au point de modèles du rendement des capteurs acoustiques sous-marins précis. Dans la présente étude, on propose une suite progressive de capacités de prédiction environnementale qu'il faudrait mettre en œuvre à l'intérieur du système de traitement sonar PLEIADES. En fin de compte, cela donne une capacité perfectionnée de modélisation du rendement du sonar qui est intégrée au traitement du sonar et aux fonctions automatisées de détection et de classification. Recherches futures : La présente étude a été complétée à la fin de Depuis, on a mené des travaux pour intégrer la capacité de rendement du sonar GSR de base au système PLEIADES. En iv DRDC Atlantic CR

9 particulier, la capacité intégrée a été mise en œuvre à bord du NCSM Ottawa durant les exercices Westploy et RIMPAC, au milieu de l'année La poursuite du perfectionnement des capacités de modélisation environnementales à RDDC Atlantique se fait dans le cadre du projet 11ch, qui porte sur l'évaluation du rendement des sonars dans des conditions réalistes. DRDC Atlantic CR v

10 This page intentionally left blank. vi DRDC Atlantic CR

11 Table of contents Abstract..... i Résumé i Executive summary... iii Sommaire... iv Table of contents... vii List of figures... viii 1 Overview Project Background Scope Document Organization Basic Environmental Analysis Package Advantages of the Basic Environmental Analysis Package Effort to Implement Basic Package for MSTDCL Intermediate Environmental Analysis Package Advantages of the Intermediate Environmental Analysis Package Effort to Implement Intermediate Package for MSTDCL Advanced Environmental Package Advantages of the Advanced Environmental Analysis Package Effort to Implement Advanced Package for MSTDCL Advantages to MSTDCL of Including an Environmental Analysis Package Conclusions and Recommendations References List of acronyms and initialisms Distribution list DRDC Atlantic CR vii

12 List of figures Figure 1. Chart Display with Sensor/Track Overlay... 3 Figure 2. Environmental Analysis Worksheet with Data Shown... 4 Figure 3. The Data-flow Structure for the NUW Developed Environmental Analysis Package... 5 Figure 4. Temperature Profile Editor... 6 Figure 5. Propagation Ray Path Display... 7 Figure 6. Transmission Loss vs. Range and Depth Display... 7 Figure 7. Chart Display with Detection Range Overlay... 8 Figure 8. Environmental Analysis Functionality... 9 Figure 9. Combined Ray-Plot and Propagation Loss Display Figure 10. Advanced Environmental Analysis Package viii DRDC Atlantic CR

13 1 Overview This work was performed by General Dynamics Canada (GD Canada) to investigate the benefits and costs associated with incorporating integrated environmental analysis as part of an automated threat detection system within the PLEIADES sonar system. This work was performed under the MSTDCL TDP contract, W Project Background For the Multi-Sensor Torpedo Detection, Classification, and Localization (MSTDCL) TDP, Defence Research & Development Canada (DRDC) Atlantic and GD Canada are developing an integrated, multi-sensor system for automatically detecting and classifying torpedoes. The TDP includes a number of sea trials where the prototype sonar processing system (named PLEIADES) is deployed on Halifax-class Frigates. Torpedo detection, by necessity, relies heavily on underwater sensors (sonars) and their processing systems. Acoustic detection, classification, and localization, whether performed automatically or not, require some understanding of the capabilities of each sensor. Therefore an ability to predict sonar performance is essential. Unfortunately, the detection performance of these systems can be strongly dependent on frequency, direction, depth, and environmental conditions. This is especially true in littoral waters. Given these considerations as a starting point, it is natural to ask whether a fast, accurate acoustic prediction capability could be used to improve the performance of the MSTDCL system. The purpose of this study is to discuss the advantages of incorporating an environmental analysis package as an integral part of this system. 1.2 Scope The guidelines for the study were to assume a starting point based on the environment package being developed by the NUW TDP. The NUW package (see Ref. [1]) is therefore defined as the basic environmental analysis package, although it provides a significant and highly useful capability. Those capabilities are first defined as a reference for further discussions. We then define an intermediate package that provides more information to the operators and command, proposes to automate some data input requirements, and adds a low frequency (<1 khz) acoustic model for improved predictions in this region of the spectrum. Finally we considered the variability of the littoral region of operations and propose an advanced environmental analysis package. The capabilities in the advanced package should provide improved sonar performance prediction accuracy, or at least provide the operational team with good insight as to the variability in the performance predictions that could exist. Another important feature for any seagoing environmental analysis package is to provide results in a timely and efficient manner. The limiting factor in providing timely results is usually the underlying acoustic propagation model. The NUW package has selected the Bellhop acoustic model, which has been shown to provide good acoustic predictions in a very reasonable DRDC Atlantic CR

14 timeframe. However as currently configured this package does require some operator experience and attention to produce the desired output products. The proposed advanced package makes every attempt to reduce the amount of operator interaction and to provide even more timely updates of the predictions. Features such as oceanographic and seabed modeling that produce updated acoustic model inputs can reduce operator interaction and improve the accuracy of the results. Such a system can obtain the necessary inputs from databases, in situ measurements and ambient noise measurements. The goal with this advanced package is to provide guidance to the tactical team on the variability that can be experienced due to the range-dependent environment and at the same time, indicate sonar deployments that try to minimize this effect. 1.3 Document Organization The report begins with three sections describing the three proposed levels of environmental package and how they would interface to the MSTDCL system. In each case, the advantages of the different packages are compared. The report also discusses the amount of effort required to incorporate each package into MSTDCL. The report then describes the general advantages of environmental analysis to MSTDCL, and concludes with a set of study recommendations. 2 DRDC Atlantic CR

15 2 Basic Environmental Analysis Package We selected the package currently being developed by the NUW TDP at DRDC Atlantic as the starting point for our analysis of the benefits of environmental modeling for MSTDCL. This section will describe the high-level capabilities of that package. For further insight into the NUW package reference [1] should be consulted. While labelled as the basic analysis package, this tool is really a significant advance over many other products readily available. For example, the operator can choose to view the results of the modeling in many ways through a variety of displays. A typical operator display from this product is shown in Figure 1 below. The higher level displays are chart-based with overlays of sonar positions and predicted detection ranges if desired by the operator. Figure 1. Chart Display with Sensor/Track Overlay DRDC Atlantic CR

16 The operator interfaces with this package by editing a work sheet. An example of this worksheet is shown, with some of the typical data values used, in Figure 2 below. The setup of the work sheet inputs does require a certain amount of oceanographic expertise and experience, but this is the case with most environmental analysis packages. Figure 2. Environmental Analysis Worksheet with Data Shown The operator interaction is quite extensive. This is not significant when using the package as a planning tool, but can be quite a strain on the operator resources when engaged in an operational scenario. Of course a number of the inputs will be constant for the scenario at hand, but with other inputs (such as the temperature profile) more detailed actions are required. In Figure 3 below we show a general flow diagram for this basic environmental analysis package, including the operator interaction that must occur in order to develop certain data inputs for the model. 4 DRDC Atlantic CR

17 BT SVP Database Bathy Extrapolation Interpretation Operator Interface Seabed Databases Bottom Loss Parameters BLUG Environmental Analysis Manager Service Bellhop (DRDC-A) Sensor Noise Data Wenz Curves Acoustic Detection Performance Prediction; SQR-19, Sonobuoy AN/SQS-510 (Active & Passive) Databases Sonar Source Target Performance Predictions Figure 3. The Data-flow Structure for the NUW Developed Environmental Analysis Package The fundamental acoustic model behind this analysis package is the BellhopDRDC model [2]. This BellhopDRDC Gaussian Beam ray model has had some additions and refinements added as part of the NUW project and this improved version is being embedded in the Environmental Analysis tool package. The Bellhop model is well suited to both the NUW and MSTDCL applications since it offers timely, accurate results over most of the frequency bands of interest and accommodates range-dependent bathymetry. To begin an environmental analysis, the user must select a location and provide a temperature profile. The tool allows temperature profile inputs from measured BT data or historic profiles that in turn can be blended with the observation if required. Because the BT data must be manually edited, the package is prone to error in a tactical environment unless an experienced acoustic operator is available to periodically carry out this important input task. It is more suited to use as an off-line pre-mission guide. The selection of the measured points to employ and the merging with other data can affect both the speed of operation and the accuracy of the predictions. An example of the profile editing display and functionality is shown below in Figure 4. The operations allowed with this editor are: Add temperature point; Move temperature point; DRDC Atlantic CR

18 Save profile; Load profile; Truncate down; Truncate up; and Down sample temperature profile. Figure 4. Temperature Profile Editor There are background software tools provided to assist the operator in the data point selection such as the broken pipe algorithm which provides a guide for the down sampling process. The NUW environmental package contains databases (or spaces to insert databases) for historic temperature and sound velocity profiles, bathymetry, bottom province and bottom loss data. It allows access to the bathymetry database via latitude/longitude co-ordinates and then along a specific operator selected bearing or look direction. The acoustic model outputs are propagation loss and ray plots for the specified conditions. Examples of the currently provided displays for the propagation paths and propagation loss results for the direction selected are shown in Figures 5 and 6 below. 6 DRDC Atlantic CR

19 Figure 5. Propagation Ray Path Display Figure 6. Transmission Loss vs. Range and Depth Display The propagation loss predictions are converted to detection predictions using noise data, sensor performance inputs and target characteristics. The sensor and target parameters are taken from a stored database of known values, along with operator-entered values for the receiver depth. Noise levels can be taken from through-the-sensor measurements, historic data or Wenz curves for ambient noise. In the active case, the NUW package has attempted to provide a reverberation model for pre-mission planning purposes, or it can use through-the-sensor measurements to obtain these values. The detection prediction results can be shown to the operator on a chart display, as overlays on the sensor positions. While the calculations are for one receiver depth the full field of target depths is computed and a 2-D plot of this data is planned. One of the strongest features of this NUW Environmental Analysis package is the chart display. It allows the operator to work at a high level to view the expected detection ranges and the deployment of the sensors. It gives the package an excellent planning tool capability as well as an operational overview of how well the deployed resources are functioning against the specified targets. The areas of coverage for each sensor are clearly visible to the operator. When variable detection ranges exist within the water column, they are displayed as maxima and minima for the receiver depth selected. These are indicated by the two colour overlays as seen in Figure 7 below. DRDC Atlantic CR

20 Figure 7. Chart Display with Detection Range Overlay There were several other features planned for the NUW Environmental Modelling Manager (EMM) package, for example optimum sensor placement in the water column, optimum ping selection, etc. which were not completed. These Functions, along with the absence of certain database information, are some of the components of the NUW EMM package that will need to be addressed in the MSTDCL Intermediate level environment package if full advantage of this analysis capability is to be achieved. An overview of the NUW package functionality is shown in Figure 8 below. 8 DRDC Atlantic CR

21 Source Characterization magnetic Characterization Ambient Noise estimation Ambient Noise assesment Signal Detection estimation Sensor Characterization surface Characterization Surface Scattering estimation Surface Scattering assessment Echo Detection estimation Target Characterization Bottom Characterization Bottom Scattering estimation Bottom Scattering assessment Sensor Placement Optimization Salinity characterization Transmission loss estimation Transmission loss assessment Source Placement optimization Temperature Characterization Reverberation level estimation Reverberation level assessment Ping selection Optimization Sound Speed Characterization Propagation path estimation Data management data management visualization Environmental management Environmental Visualization EA application management EA application visualization Not implemented Figure 8. Environmental Analysis Functionality There are three assessment blocks which are shown as not completed, and these involve the further analysis of the through-the-sensor measurements to provide the additional information that is required. Some additional processing of the received signals will be required to complete the information gathering intended by the functionality illustrated. For example, the surface scattering assessment provides a measure of the frequency spreading, which in turn will be used to provide guidance to the operators on the minimum target speed for good Continuous Wave (CW) detections. In some cases this assessment information may be required as part of the inputs for the optimization calculations and for the echo detection estimation. The associator, tracker and classifier can have improved acoustic path prediction data, better range estimates for bounds on DRDC Atlantic CR

22 possible tracker solutions, and improved target parameter estimates. These will be discussed further below. 2.1 Advantages of the Basic Environmental Analysis Package While other environmental analysis packages do exist and could be purchased, they tend to be very much stand-alone sonar performance prediction packages. The NUW package has evolved to include some of the more in-depth displays included in these other models, but goes well beyond this capability. For example the chart display with overlays and sensor positions gives the operator a higher-level view of the acoustical situation that is much more tactically relevant. The NUW package also includes the use of through-the-sensor measurements to update the performance predictions based on local, current environmental measurements. While some degree of operator input is required by the NUW package during the data preparation and selection of acoustic model inputs, this can be much less than that required by some of the competing packages. The NUW package was also constructed with operational employment in mind so that the time to update and achieve a new set of acoustic predictions is much faster than that of other acoustic modeling packages. The NUW package has already worked out the Calibration Factors (CF) applicable to CF sonars, i.e. the SQR-19 array, the SQS-510 sonar, and Directional Frequency and Ranging (DIFAR) sonobuoys, so that their through-the-sensor measurements can be readily incorporated into the performance predictions. This provides the MSTDCL automatic processing, associators, and trackers with a valuable second source of possible target detection ranges, as well as possible overlapping detection areas, which are important parameters in the fusion of contacts on multiple sensors to the same target. On a more pragmatic level, the technology developed for the NUW TDP is wholly owned by DRDC and was developed by the local office of GD Canada. Furthermore, it was developed using the same System Test Bed (STB) architecture that forms the basis of the PLEIADES system. All of this means that the NUW package has strong local support and can be transferred to MSTDCL with a relatively small amount of cost and effort. For these reasons the NUW approach to performance prediction and environmental modeling is recommended as the basic starting package for the MSTDCL requirement. It is worth noting, however, that GD Canada has also been developing its own environmental prediction capability. The development of the Sensor Performance and Acoustic Detection Evaluation System (SPADES) and its primary user interface, the Integrated Coverage Editor (ICE), has been funded as an internal research and development project at GD Canada, led by Dr. Gary Brooke and Steve Kilistoff in Victoria. The SPADES/ICE system was deployed on a CPF during an MSTDCL sea trial in February It provides a very efficient, advanced environmental prediction capability, similar to that provided by NUW. In fact, although it lacks some of the advantages identified above, it does include some features proposed for the intermediate environmental package as described in Section 3. Therefore, the extra cost of integrating SPADES into the STB should be balanced against the cost of upgrading. 10 DRDC Atlantic CR

23 2.2 Effort to Implement Basic Package for MSTDCL Since NUW has implemented its Environmental Manager as a set of STB components, there is minimal effort required to add this functionality to MSTDCL. The tasks required include: Adding the display tools to our Chart display; Adding the necessary operator inputs as Java panels (vice the current X-Windows tools); Modifying the signal processing to provide calibrated outputs for through-the-sensor noise estimates (the NUW TDP has defined the algorithm); Completing the Mk8 interface; and Configuration, integration, and testing. In total approximately 40 days of effort is estimated to be required for these tasks. DRDC Atlantic CR

24 3 Intermediate Environmental Analysis Package The proposed Intermediate Environmental Analysis package builds on the NUW package to provide the operator with several major refinements and improvements. These are: A more detailed evaluation of the environment through more extensive use of databases, automation of the environmental assessments, and more comprehensive combined displays; Completion of the sensor placement optimization to assist operators and command in the optimum sensor placement; Including a Parabolic Equation (PE) acoustic model in the environmental package for frequencies below 1 khz; Extending the system to provide continuous updates to the detection predictions based on the current geographic location; Automation of model parameter selection; Automation of Sound Velocity Profile (SVP) data preparation; and Completion of the database population. An example of more detailed environmental assessment is shown in Figure 9 below, where the ray-plot and the full field propagation loss have been combined into one display. While based on a single Bathythermograph (BT) assumed constant over range, the model run blends the measured and historical BT data to accommodate the variable bathymetry automatically. 12 DRDC Atlantic CR

25 Figure 9. Combined Ray-Plot and Propagation Loss Display A second feature of the intermediate package is provision of outputs to assist command and the sonar operators in a coarse estimate of the optimum placement of the receivers; for example, where to deploy sonobuoys, and what depths to choose for the sonobuoy hydrophone. In the case of the towed array the tool would allow the operator to input the desired speed of advance and cable scope range. This will be converted to a range of depths available, and the model runs will be used to select a couple of options. These will guide the placement of this important sensor for best water column coverage. Most of the calculations are already done for this type of analysis within the model. The intermediate package will need an operator interface to perform the whatif analysis. The Bellhop model uses ray-path predictions that have had some error problems, especially below 1 khz. In the intermediate level package it is proposed to include the ability to run a second model suited to low-frequency passive propagation problems. A reasonable choice would be a PE model. These models are reasonably efficient for low frequency shallow water scenarios, and DRDC has ready access to a well-established PE model (PECan). In those spectral areas below 1 khz, where the PE results and the Bellhop results differ by more than a few decibels, then this package would select and use the PE results in the computations for system performance and coverage indications. DRDC Atlantic CR

26 Another proposed change for the intermediate level package would be to have the environmental analysis system running in the background using the current location and through-the-sensor noise measurements to update the detection performance. This would alert the system operators to any look directions or sensor types whose performance may have changed from the time of the initial environment analysis runs. We further propose to automate the calculations for all sensors in the task group, which generally includes the SQS-510, AQAP-3 (helicopter dipping sonar), Canadian Towed Array Sonar System (CANTASS) and sonobuoys. We would also automate the calculation of counter detection ranges for our own signatures at the current speed, and for the SQS-510 hull-mounted sonar. There are no significant technical hurdles preventing real-time operation of the NUW package. One final improvement that has high importance in the intermediate level package is automating the preparation of the SVP data and the geoacoustic data for input into the acoustic model. Currently in the basic package this requires operator intensive effort, but the knowledge exists to automatically configure this data. There already exist bathy launchers with computer interfaces for data transfer, and MSTDCL has partially implemented an interface to the Mk.8 Bathythermograph on the Halifax-class ships. We require smoothing of the data received, a decision on the actual water depth, and reconciliation of this measured data with historical database data for extensions where the depth profile has deeper water than the point at which the measurement was taken; i.e. a down slope look direction. Also we require merging of the BT data with satellite or ship s surface measured surface temperature to more accurately represent the first 10 m of profile data. It is important to note that these latter improvements are not aimed at improving the fidelity of the propagation model (Bellhop), which appears to be adequate for most scenarios. In the basic package, however, the accuracy of the model results is limited by inaccurate and out-of-date environmental data-base parameters. Many of the upgrades proposed for the intermediate package, by automating measurements and data synthesis, would improve the reliability of the environmental specification, and by extension would result in more accurate and reliable acoustic performance predictions. The operators and command would have constantly up-to-date information with very low operator involvement, which is a significant advantage for the intermediate level environmental analysis package. 3.1 Advantages of the Intermediate Environmental Analysis Package The intermediate level of acoustic environmental analysis package represents a significant engineering upgrade to the basic package provided by NUW. In the intermediate level analysis package all the advantages attributed to the basic package currently being developed by the NUW TDP are applicable, along with the added advantages of: Reduced operator intervention; Completed database population; Completed optimization for ping selection and sensor placement; 14 DRDC Atlantic CR

27 More frequent updates via automatic model parameterization; More accurate predictions for frequencies below 1 khz; and Improved operator/command communications via the updated displays and additional outputs provided. In the case of MSTDCL operations in littoral waters, where the acoustic environment can be changing rapidly, it is a significant advantage to have more frequent updates. At a minimum the operator is alerted to the changes in the sonar coverage and has the potential to adjust the sensor depth in the case of the towed array, deploy additional sonobuoys to the new optimum depths, and to inform command of the changing detection coverage available from the deployed sensors. The more frequent updates can only be considered viable in the operational context if the process of editing the bathy information is automated. This involves potentially changing the bathy launcher on the frigates to include an updated multi-chamber launcher and the elimination of operator intervention in doing the editing and preparing the profile for input to the acoustic model. This automation provides for more timely updates of the environmental picture and is viewed as a significant advancement for this MSTDCL intermediate level package. The inclusion of the PE acoustic model in this package for frequencies below 1 khz improves the performance prediction accuracy in this region of the acoustic spectrum for environments where ray path analysis has been shown to have some weaknesses. The addition of this acoustic model to support the Bellhop model is a small step that NUW may implement before MSTDCL imports this functionality. The addition of this low frequency model is a prediction accuracy advantage for MSTDCL since most signals of interest for the CANTASS and sonobuoys are below 1 khz. This intermediate level environmental package will provide improved inputs for the MSTDCL automatic processing components, the associator and the tracker. With the basic package the accuracy of the detection range estimates which may be based on the SVP from a few hours ago can at best provide a generalized range of the day sort of estimate. Even though the bathymetry and noise backgrounds would be updated in the basic package, the rapidly changing environments of the littoral region can often require more frequent SVP updates which this intermediate package proposes to make available. This provides the associator and tracker processing more current and reliable estimates on which to base the solutions. One additional advantage for this intermediate level package is the improved communication options provided by the additional displays and processing. This would provide the PLEIADES operator the additional knowledge on how to deploy the sensors to achieve the best performance against a particular threat. In some environments one sensor placement may not be as viable as another, and the consequence of this may be reduced detection ranges regardless of the sensor depth selected. However, having an operator informed of the sonar environment, through the additional displays of the detection coverage possible with differing deployments, increases the effectiveness of the PLEIADES system. DRDC Atlantic CR

28 3.2 Effort to Implement Intermediate Package for MSTDCL This section contains very rough estimates of the order of magnitude of effort that would be required to implement the Intermediate Environmental Package. As a first step, the Basic Environmental Package must be implemented. If DRDC is interested in pursing these capabilities, then GD Canada recommends performing a design investigation to better define the scope, solution, and projected costs. GD Canada views this effort as a natural extension of this study. If directed by DRDC, GD Canada will propose how to conduct such an investigation within Option Phase 3 of the MSTDCL contract. The tasks required and their rough order of magnitude estimates are: A more detailed evaluation of the environment (new display formats): days; Coarse outputs to assist command in the optimum sensor placement: days; Including a PE acoustic model for frequencies below 1 khz: days; Extend the system to provide continuous updates to the detection predictions based on measured noise and the current geographic location; days (includes automating performance of all models of interest to the task group); Add sensor coverage inputs to the tracker, associator and classifier to provide more reliable system outputs: days; and Automate SVP data preparation: days. The total estimate is on the order of 200 days, +/- 50. Clearly this is a significant investment; however, it is the type of functionality that is required to significantly boost the operational performance of our sonar systems in realistic at-sea environments. As an alternative approach, note that GD Canada s SPADES tool already provides some of the Intermediate capabilities described above. Therefore, the cost of upgrading from the Basic package to the Intermediate package would be reduced. This must be balanced against the higher initial cost of integrating SPADES with the STB. 16 DRDC Atlantic CR

29 4 Advanced Environmental Package In the basic and intermediate environmental packages we assumed databases of sufficient accuracy were available so that the acoustic model was provided with inputs that would describe the range independent environment. In deep water the resulting predictions would be sufficiently accurate over long ranges. In the littoral regions we cannot make that assumption. The main thrust of the proposed advanced environmental package for MSTDCL is to enable more accurate and detailed environmental assessment, using in situ and through-the-sonar measurements. In littoral environments one SVP will not apply over all ranges and areas for which predictions are required. Another difficulty is that under typical situations the Frigates usually only acquire a few SVP per day. In a littoral environment the SVP may change on spatial scales of a few miles, so this level of sampling is inadequate. In addition the bottom parameter databases are, in many cases, actually holdovers from deepwater applications. One of the current geoacoustic databases, the Bottom Loss Upgrade (BLUG) database, is an extrapolation from deep-water provinces to shallow water bottom types and is rarely based on local acoustic measurements. For these reasons and the complexity of littoral waters, it is not uncommon for model predictions of propagation loss, and hence detection range estimates, to be in disagreement with measurements by 5 10 db or more. Computing the full range-dependent environment is desirable, but this is a challenging problem. For the ranges of interest to the MSTDCL project it is still very worthwhile to consider an advanced acoustic environment analysis package that accurately handles a range-independent SVP and geoacoustic properties along with the range dependent bathymetry. This is similar to the approach used in the basic and intermediate packages. However in the advanced package we propose automated tools that can update the model inputs based on sensor measurements, producing an ensemble average for the SVP and geoacoustic inputs, more representative of the true environmental complexity. That is, the ensemble average is representative of the range-dependent environment and accurately reflects the expected variability that could exist in the environmental parameters over the ranges of interest. In an advanced environmental analysis package the goal is to reduce disagreements between the performance prediction and measurements, where possible through in situ geoacoustic inversions (single sensor), environmental measurements with tomography (SVPs), and ambient noise measurements. It is assumed that an initial mean environment that is close to reality can be estimated from databases and measurements. These include the SVP and a geoacoustic profile that includes sediment sound speed, density and attenuation versus depth. Multi-path propagation in a shallow water environment leads to coherent structure in the sound field both spatially and temporally. The nature of this coherent structure is strongly dependent on the geoacoustic parameters of the sediment and is the physical basis for the modeling approach to the extraction of the geoacoustic parameters. The goal of the advanced package is to predict the actual range dependence for both the water column and the seabed properties. Two methods for generating and handling the inputs are addressed in this advanced environmental package. The first is an interpolating scheme for the SVP and the geoacoustic parameters to be used in the acoustic model. We will need to impose a time constraint to ensure this converges in time to provide useful outputs. If it does not, or if we DRDC Atlantic CR

30 cannot use this approach because the range-dependent SVP is not known, then a second statistical approach can be used where a mean profile is perturbed based on observation. This allows the acoustical model, and hence the operator, to see a meaningful range of performance estimates which can then be used to guide the deployment of resources and the interpretation of sonar results. The approach is comprised of five basic steps: 1. An analytical estimate of the critical environmental parameters is performed. Beginning with a mean SVP it can be predicted which of the environmental parameters would have the most influence on the propagation paths at hand. 2. A measurement of these critical environmental parameters is performed, along with statistical uncertainties. 3. Generate an ensemble of possible environments. Here we consider a range dependent environment with one possible SVP perturbed slightly, to describe variability over the entire path length. Similarly for the geoacoustic features to be passed to the acoustic model, we consider an ensemble of possible parameters. 4. Compute the transmission loss and ray paths for an ensemble of environmental parameters. 5. Communicate the acoustic uncertainty to the operator in a meaningful way. The proposed advanced environmental analysis package is illustrated in block diagram form in Figure 10 below. One attribute of this advanced system is that the mean SVP will be used to guide the selection of the environmental parameters that have the most effect on the model results. For example, it is well known that steep propagating acoustic rays (high normal modes) refract little and are insensitive to details of the sound speed. We also know that sound channel or trapped acoustic energy as well as surface reflected-bottom refracted rays are insensitive to geoacoustic parameters. 18 DRDC Atlantic CR

31 Bathy or SVP or CYD Data CCS Interface Ocean Database Remote Sensing Data OCEANOGRAPHIC MODELLING Operator Interface Insitu Measurements Acoustic Measurements Acoustic Tomography ECPINS Oceanographic Databases SEABED MODELLING Environmental Analysis Manager Services Acoustic Models Bellhop 1kHz-20kHz PE 10Hz-1 khz Insitu Measurements Acoustic Measurements Geoacoustic Inversion Sensor Noise Noise Data Wenz Curves Databases Sonar Sources Acoustic Performance Predictions: AN/SQS-510 Active/Passive: SQR-19, Sonobuoy Targets Figure 10. Advanced Environmental Analysis Package We must consider the source and receiver depths when dealing with environmental parameter classifications that are derived from the sound velocity profile. Acoustic field measurements made at different depths will be sensitive to different environmental parameters. For example we can classify paths as downward refracting, upward refracting, surface duct, sound channel, etc. Analysis in Ref[2] indicates how the model can use these classifications. For example, downward refracting rays or modes (acoustic rays/modes with phase speed at the source that are less than the surface sound speed) will be refracted away from the surface and can be considered examples of surface-reflected-bottom reflected paths. These paths are sensitive to the SVP and to internal wave activity in the mixed surface layer. The energy will also interact with the bottom though with lower grazing angles. For hard sediments with high sound speed ratios and high critical angles all the bottom interaction will be sub-critical and propagation will be sensitive to the sediment attenuation but not the sediment sound speed. Similar analysis can be derived at DRDC Atlantic CR

32 other source depths for the same SVP where the phase speed is greater than the surface sound speed and these produce surface reflected-bottom reflected paths. All this energy will be at relatively high angles and will therefore refract little as they propagate. This analysis is meant to generate an appreciation of the variability that may be encountered in the acoustic model outputs based on the level of confidence that can be placed on the database type and the sampling resolution. This analysis in turn will lead to an automatic input and update system for the parameters required by the acoustic model. This automatic input system will track the ship movement and update the operator performance prediction displays on a periodic basis. The advanced acoustic environmental analysis package relies on the ability of the model to incorporate local measurements from passing ships of opportunity. When these contacts exist, we can develop a scissorgram, a measure of the received acoustic power level as a function of radar range. Through analysis we can invert these power levels to provide more reliable estimates of the local environmental parameters, the geo-acoustic parameters and the modified SVP, than can be supplied to the acoustic model, by databases alone. When the input parameters have been derived from such measurements the results of the performance predictions can be presented in a different color on the predictions chart, to clearly indicate to the operator that these results are based on reliable local data measurements. In this advanced environmental analysis package an effective parameterization of the environment is sought which permits the accurate prediction of the features of the acoustic field that are relevant for the sonars involved and the range dependent environment they are working in. A second aspect of this advanced package is rapid computation for timely availability of the desired sonar performance predictions. This means that the generally very large geoacoustic parameter search space is significantly reduced by using a simple geoacoustic model with a single layer, single grain size sediment, overlaying a hard acoustic half-space. This model is based on the work of Hamilton [4] and Buchmann [5]. In this advanced model the uncertainty of the input data will depend on the area of operations, the databases available for that area, and the operational conditions encountered. When we encounter contacts of opportunity that allow us to estimate the environmental parameters, we can compare the results to the uncertainties. As we develop a statistically significant set of measurements, we can better quantify the uncertainty, which leads to better ensembles and better bounds on the possible environmental conditions. 4.1 Advantages of the Advanced Environmental Analysis Package In the previous two levels of acoustic environmental analysis and prediction packages a number of the input parameters were obtained from either historical databases or from generalized geoacoustic tables. However variability in the coastal ocean environment spans a wide range of spatial and temporal scales and these can lead to poor acoustic predictions. The key advantage of this advanced acoustic environmental analysis package is that it addresses the uncertainty which may be present in the input environmental parameters and alerts the operator to the potential variability which may exist in the results obtained from the performance prediction tools due to these environmental uncertainties. 20 DRDC Atlantic CR